The short answer

Why you lose

Every football bet at a bookmaker's price carries the bookmaker's margin, so its expected value is negative before kick-off. No staking system — Fibonacci, Martingale, chasing the draw — changes that sign: it only changes the shape of the loss, from many small ones into one large one that arrives later. No pattern in past results changes it either: a long run without a draw does not make a draw due. The only thing that wins over distance is a probability more accurate than the price, and on 545,000 matches our own models do not manage it. This section shows each of those claims on the data.

Three facts from the data

N = the number of matches a team had gone without a draw before the match in question.

Flat-stake return at N = 10–14 -3.71% One unit on the draw in every team-match after a 10–14-match run without a draw, at Pinnacle's closing price. 12,122 priced team-matches in the historical market index. Computed 2026-09-30. This is the market page’s flat-stake comparison, not a Fibonacci staking return.

Is a draw due? 25.7% vs 17.8% Draw rate after N = 0 vs N ≥ 31, for teams with 100+ earlier matches in the historical draw-due index. Team-season dispersion ratio: 0.996 (1.0 means independent events). Computed 2026-09-30.

Matches indexed 545,258 Played matches across the whole archive, using the same count as the footer. Coverage varies by competition and era.

In this section

Is a draw due?
Luck or skill
Bank, lines and the market (requires an account)
Why chasing doesn't work
The systems sandbox
The bookmaker's margin (coming)
Selective memory (coming)
Can a model beat the market? (coming)

Calculators

This site takes no bookmaker advertising and sells no system.